Vehicle Minor Collision Detection Using Telematics and Environmental Data

Road safety affects everyone, not just Geotab customers. With several years of driving and environmental data collected from over 2 million connected vehicles, there is a great opportunity to leverage big data and machine learning to establish a minor collision detection system. On top of driving data and environmental data, it also contains machine diagnostic data. All these datasets are hypothesized to contain hidden features that are highly correlated to minor collisions. Since minor collisions are harder to identify, high frequency event data will be required. Telematics data is typically heavily compressed, however Geotab has a system that is capable of capturing higher frequency sampled data that will make this possible. The objective is to use this to detect minor collision occurrences through a combination of machine learning approaches which fall under the umbrella of supervised and semi-supervised learning.

Faculty Supervisor:

Andrei Badescu;Sheldon Lin

Student:

Partner:

Geotab Inc

Discipline:

Computer science

Sector:

Information and cultural industries; Professional, scientific and technical services; Transportation and warehousing

University:

University of Toronto

Program: